matvogel
Member
Hello Folks!
After using Olympus fourthirds for a while I bought a used R-D1 last week. I already discovered that my R-D1 also has the almost mandatory rangefinder disalignment which has very often reported and this will fixed soon by my local repair man. Beside of this I made some compararison shots with my fourthirds based Olympus 510. The comparison shows very disappointing results to me.
R-D1 was equipped with two lenses:
1. VC Nokton 50/1.5
2. Zeiss Sonnar 50 /1.5
The Olympus was equipped with the Zuiko 14-54 zoom, set at about 40mm
I used a stable tripod. All lenses where set at infinity and f/8
Exposure time was 1/30 with the R-D1 and 1/20 with the Oly.
For fair comparison I used Camera Raw for all three files. The only manipulation was a slight rescaling of the Olympus file because it has a 10 MP sensor instead of the R-D1's 6MP.
I include two different crops of all three original Raw files. They where directly stored as Tiffs, no JPEG compression in between. Only for uploading here I converted them to JPGs at the highest qulality level.
Do you guys have an idea why the Olympus file is so much sharper than the VC and the Zeiss glass? I would expect the opposite behavior because the fourhirds sensor is smaller than the APS-C of the R-D1. And the image quality of the R-D1 was always mentioned in a very good context.
I'm looking forward in your ideas and comments.
Regards
Matthias
Olympus Crop1
Nokton Crop1
Sonnar Crop1
Olympus Crop2
Nokton 50 Crop2
Sonnar Crop2
After using Olympus fourthirds for a while I bought a used R-D1 last week. I already discovered that my R-D1 also has the almost mandatory rangefinder disalignment which has very often reported and this will fixed soon by my local repair man. Beside of this I made some compararison shots with my fourthirds based Olympus 510. The comparison shows very disappointing results to me.
R-D1 was equipped with two lenses:
1. VC Nokton 50/1.5
2. Zeiss Sonnar 50 /1.5
The Olympus was equipped with the Zuiko 14-54 zoom, set at about 40mm
I used a stable tripod. All lenses where set at infinity and f/8
Exposure time was 1/30 with the R-D1 and 1/20 with the Oly.
For fair comparison I used Camera Raw for all three files. The only manipulation was a slight rescaling of the Olympus file because it has a 10 MP sensor instead of the R-D1's 6MP.
I include two different crops of all three original Raw files. They where directly stored as Tiffs, no JPEG compression in between. Only for uploading here I converted them to JPGs at the highest qulality level.
Do you guys have an idea why the Olympus file is so much sharper than the VC and the Zeiss glass? I would expect the opposite behavior because the fourhirds sensor is smaller than the APS-C of the R-D1. And the image quality of the R-D1 was always mentioned in a very good context.
I'm looking forward in your ideas and comments.
Regards
Matthias
Jinky
Member
The E-510 and E-410 are 2 of the sharpest cameras Olympus have made in recent years, allied to one of their best lenses that would explain the sharpness of the crops above.
ampguy
Veteran
Yes, I have an idea. The Olympus images are in focus. The RD1's are off.
matvogel
Member
Hm, this was also my first idea because I still have the slight rangefinder missalignement. But the subject is at least 25m away from the camera. I set focus on infinity and aperture on f/8. Should be sharp...
LCT
ex-newbie
Sounds like you have a focus problem with your R-D1 (sharper backgrounds).
pagpow
Well-known
djonesii
Well-known
Too make a really full assessment, we need to know where the crops come from. Although, I would expect by F8 it does not make that much difference. Extreme corners might have something to do with it.
matvogel
Member
matvogel
Member
It seems to be a focussing issue caused by the rangefinder missalignement :bang:. I made another shot at much worse light conditions only with the Nokton and set it at hyperfocal distance. And here are the crops:
Both now look sharper indeed! Now in a side by side comparison with the Oly crops the new Nokton shots are now sharper!
Thanks to all of you!
Both now look sharper indeed! Now in a side by side comparison with the Oly crops the new Nokton shots are now sharper!
Thanks to all of you!
RichC
Well-known
Looks like you need to visit my R-D1 rangefinder adjustment page! <grin>
But, first, do make sure your Nokton isn't out - check focus using another lens that you're confident focuses accurately... more likely to be the R-D1 but you really don't want to adjust your camera to a misfocusing lens!
But, first, do make sure your Nokton isn't out - check focus using another lens that you're confident focuses accurately... more likely to be the R-D1 but you really don't want to adjust your camera to a misfocusing lens!
ampguy
Veteran
yet more empirical evidence that DOF scales of Leica and Leica based lenses are accurate on APS-C and H sensor based bodies in addition to legacy sizes.
semilog
curmudgeonly optimist
The RD-1 images might or might not be improved, but I'm not a bit surprised by the excellence of the Olympus.
People (especially those who have not used the system extensively) consistently underestimate the resolution that the Olympus 4/3 system is capable of. The 14-54 is an absolutely terrific lens, and the 12-60 SWD is even a bit better, providing Nyquist-limited resolution (48+ lp/mm) almost into the corners, except at the extreme ends of the zoom range, where it still does pretty darned well. Even with smallish pixels, many of the 4/3 lenses are so good that even at 12 megapixels these systems are sensor-limited.
Now, note that we're talking about Nyquist-limited resolution on a 10 or 12 megapixel sensor, vs. the R-D1's 6 megapixels. Even with perfect focus and a diffraction-limited lens, the RD-1 cannot do better than a Nyquist-limited lens on a 10 megapixel sensor.
The only way to substantially improve on the Olympus's resolution with the 14-54 or 12-60 is to go with a high pixel-count full frame sensor like that on the 5DII, or (for a far smaller improvement, perhaps 20% linear) go with a high-density APS-C sensor like that on the Canon 7D, and use only the very best lenses available. With the 7D, the lens would have to be Nyquist-limited to see a difference, and on an enlarged print a difference might not be visible even then. For a meaningful difference you need a full-frame sensor (5DII, M9, etc.).
By the way, since you were using a tripod, did you turn off the E-510's IS? If not, doing so would likely give an even better result, especially combined with MLU and remote release.
People (especially those who have not used the system extensively) consistently underestimate the resolution that the Olympus 4/3 system is capable of. The 14-54 is an absolutely terrific lens, and the 12-60 SWD is even a bit better, providing Nyquist-limited resolution (48+ lp/mm) almost into the corners, except at the extreme ends of the zoom range, where it still does pretty darned well. Even with smallish pixels, many of the 4/3 lenses are so good that even at 12 megapixels these systems are sensor-limited.
Now, note that we're talking about Nyquist-limited resolution on a 10 or 12 megapixel sensor, vs. the R-D1's 6 megapixels. Even with perfect focus and a diffraction-limited lens, the RD-1 cannot do better than a Nyquist-limited lens on a 10 megapixel sensor.
The only way to substantially improve on the Olympus's resolution with the 14-54 or 12-60 is to go with a high pixel-count full frame sensor like that on the 5DII, or (for a far smaller improvement, perhaps 20% linear) go with a high-density APS-C sensor like that on the Canon 7D, and use only the very best lenses available. With the 7D, the lens would have to be Nyquist-limited to see a difference, and on an enlarged print a difference might not be visible even then. For a meaningful difference you need a full-frame sensor (5DII, M9, etc.).
By the way, since you were using a tripod, did you turn off the E-510's IS? If not, doing so would likely give an even better result, especially combined with MLU and remote release.
Last edited:
matvogel
Member
I already checked your site. Great work!!! I'm still thinking about if I can do the alignment by myself...
Maybe you can aswer me this question: at my R-D1 the focussing patch is tilted a few degree to the left and is not in parallel to the framing lines. Do you have an idea what could be the problem?
I tested close focussing accuracy using a text page tagged on a sloping board. The rangefinder is indeed misalligned. What I figured out is that the focussing patch is also vertically misaligned- almost impossible to focus on a text page. On this test page I could only focus by drawing also a vertical line. With my Zeiss Ikon no problem at all!
Regarding another lens: I only have my Sonnar 50/1.5 which should perform well at f/8. Or the CV Heliar 75 - Is the Heliar known for misalignement? All other glass is too wide for such tests.
Maybe I double check the Nokton on the Zeiss Ikon tomorrow. Film development is quickly done and first pre-check on a light table, too. Will see...
Cheers
Matthias
RichC
Well-known
Yes - poor quality control during manufacture, or the camera has had a knock (not necessarily a hard one). The Cosina-Voigtlander Bessa rangefinder (which is what the R-D1 rangefinder is based on) is not especially well designed or robust, and a common fault even in new cameras is a misaligned rangefinder (mine was brand new and way off) - Epson would have been better off letting Cosina's skilled personnel assemble the camera rather than doing it in in-house using their own less-skilled workers. Sometimes the rangefinder patch is skewed, like yours - to fix this, the top plate must be removed to allow access to the rangefinder mechanism: a camera technician familiar with rangefinders might be able to repair it, or it might need to go back to Epson.
Sadly, despite being initially keen on the R-D1 I found it not robust enough - one of the reasons I replaced it with a Leica M8.
You'll soon get used to how the skewed rangefinder patches align when focus is spot on, so personally I'd just live with it - a minor irritation.
As an aside, checking that the R-D1's rangefinder is both adjusted correctly and straight is something a buyer should always check - either to reject the camera or reduce the selling price.
I used to have a CV 75 - a bloody excellent lens on which the focus was spot on. I've owned four CV lenses (not all bought new) and they were all perfect in every way. I'd check focus with both your CV lenses - I suspect that this will confirm that the R-D1 is out of adjustment.
It's easier to adjust the R-D1 with the top plate removed - and is straightforward if fiddly. However, any competent camera repairer will be able to adjust your camera for you quickly and cheaply (and might even be able to sort out the skewed patch).
matvogel
Member
Hi Richard,
Thanks for your information. Yes, Epson for sure made wrong business decisions. No wonder why the R-D1 remained their only trial entering the camera industry - very sad! The camera is generally very well designed!
Today, I talked to my repair man and showed him your web site. He did not know the camera but repairs Cosina/Voigtländers and other rangefinders since years. He is very confident that he can well adjust the rangefinder for me. Maybe not the skewed patch, we will see. If this is not possible, I can live with that minor issue.
I was thinking quite a time about buying the M8 or the R-D1. In the end I was not sure if I like a digital rangefinder in general. Therefore I decided towards the cheaper option. If I still like it in six months I will for sure think again about other options. We will see how prices for used M8 will drop. I'm also curious if we see something new on Photokina this year.
You're right - classical mistake when buying via ebay...
Today I shot a roll of Rollei R3 at 20ASA in my Zeiss Ikon and tested at all apertures the Sonnar 1,5/50, the Nokton 50/1.5, the Color-Heliar 75/2.5 and the Nokton 35/1.4. All showed very good results as they should. The Sonnar 1,5/50 also shows the well known focus shift problem at large apertures. So I know now definitely that the rangefinder is out and the lenses are brilliant!
Thanks for all contributions in this topic!
Cheers
Matthias
Phil_F_NM
Camera hacker
I was going to say that the RF was misaligned or you have a focus shift in the lens, but these have been covered in previous replies.
Regarding getting an RD1 as a "starter" digital RF, I think it's a good choice. The cool thing about the RD1 over the M8 is that there is a more limited quantity of RD1 bodies out there and the value on them is not dropping nearly as fast as the M8/M8.2 since Epson/Cosina hasn't followed up with an updated version. So if you love the DRF and want to spring for an M8, do so after, say, 6 months to a year and you'll probably make up 90% or more of the purchase price of the RD1. They are holding their value pretty well these days, hanging right around $1000. M8 prices are kind of up & down, but more down. They regularly go for $2400 and a few have shown up below $2000.
Have fun with that great RD1, it's definitely one of a kind & has one of the best user interfaces out there.
Phil Forrest
Regarding getting an RD1 as a "starter" digital RF, I think it's a good choice. The cool thing about the RD1 over the M8 is that there is a more limited quantity of RD1 bodies out there and the value on them is not dropping nearly as fast as the M8/M8.2 since Epson/Cosina hasn't followed up with an updated version. So if you love the DRF and want to spring for an M8, do so after, say, 6 months to a year and you'll probably make up 90% or more of the purchase price of the RD1. They are holding their value pretty well these days, hanging right around $1000. M8 prices are kind of up & down, but more down. They regularly go for $2400 and a few have shown up below $2000.
Have fun with that great RD1, it's definitely one of a kind & has one of the best user interfaces out there.
Phil Forrest
sonwolf
Established
25m does not equal infinity on the CV 50mm 1.5. Even with the f/8 depth of field, the image will lose sharpness, especially in a pixel peeping test like this. From my experience, the 50mm Nokton lens favors the wider apertures and is best at f/2.8-4.0. Even at f/5.6 the image softens so f/8 is not the best aperture for a detail test.
None of this is meant as a criticism of the Olympus. It is a fine camera.
Ezzie
E. D. Russell Roberts
The R-D1 is a better camera than the 6Mpix sensor would have you believe. That and the superior M and LTM glass, even if SLR optics are catching up. Even so I am somewhat surprised that the Olympus doesn´t run away with the trophy, yes the sensor is smaller, but it does have more than 60% more pixels, and I dare say the test image was taken at base ISO or there about? So noise doesn´t come into the equation, at least not in this case?
Penceler
Established
You can't be serious? Not to sound like a Zeiss snob but we're all just wasting time & effort when an Olympus zoom (through an adaptor) yields sharper images?
John